If so, I find this reasoning unconvincing

Why?

I mostly don't agree that "the pattern is clear"—which is to say, I do take issue with saying "we do not need to imagine counterfactuals". Here is (to my mind) a salient example of a top-level comment which provides an example illustrating the point of the OP, without the need for prompting.

I think this is mostly what happens, in the absence of such prompting: if someone thinks of a useful example, they can provide it in the comments (and accrue social credit/karma for their contribution, if indeed other users found said contribution useful). Conversely, if no examples come to mind, then a mere request from some other user ("Examples?") generally will not cause sudden examples to spring into mind (and to the extent that it does, the examples in question are likely to be ad hoc, generated in a somewhat defensive frame of mind, and accordingly less useful).

And, of course, the crucial observation here is that in neither case was the request for examples useful; in the former case, the request was unnecessary, as the examples would have been provided in any case, and in the latter case, the request was useless, as it failed to elicit anything of value.

Here, I anticipate a two-pronged objection from you—one prong for each branch I have described. The first prong I anticipate is that, empirically, we do observe people providing examples when asked, and not otherwise. My response to this is that (again) this does not serve as evidence for your thesis, since we cannot observe the counterfactual worlds in which this request was/wasn't made, respectively. (I also observe that we have some evidence to the contrary, in our actual world, wherein sometimes an exhortation to provide examples is simply ignored; moreover, this occurs more often in cases where the asker appears to have put in little effort to generate examples of their own before asking.)

The second prong is that, in the case where no useful examples are elicited, this fact in itself conveys information—specifically, it conveys that the post's thesis is (apparently) difficult to substantiate, which should cause us to question its very substance. I am more sympathetic to this objection than I am to the previous—but still not very sympathetic, as there are quite often other reasons, unrelated to the defensibility of one's thesis, one might not wish to invest effort in producing such a response. In fact, I read Duncan's complaint as concerned with just this effect: not that being asked to provide examples is bad, but that the accompanying (implicit) interpretation wherein a failure to respond is interpreted as lack of ability to defend one's thesis creates an asymmetric (and undue) burden on him, the author.

That last bit in bold is, in my mind, the operative point here. Without that, even accepting everything else I said as valid and correct, you would still be able to respond, after all, that

What’s not fine is if, instead, you debit me for that comment. That would be completely backwards, and fundamentally confused about what sorts of contributions are valuable, and indeed about what the point of this website even is.

After all, even if such a comment is not particularly valuable in and of itself, it is not a net negative for discussion—and at least (arguably) sometimes positive. But with the inclusion of the bolded point, the cost-benefit analysis changes: asking for examples (without accompanying interpretive effort, much of whose use is in signaling to the author that you, the commenter, are interested in reducing the cost to them of responding) is, in this culture, not merely a "formative evaluation" or even a start to such, but a challenge to them to respond—and a timed challenge, at that. And it is not hard at all for me to see why we ought to increase the cost ("debit", as you put it) for writing minimally useful comments that (often get construed as) issuing unilateral challenges to others!